270 
BULLETIN OF THE BUREAU OF FISHERIES 
It would seem that when fat, fish might be expected to have a lesser head length than 
when thin, as the bulging of the sides of the fat fish, by increasing the angle between 
the opercle and the axis of the body, would tend to shorten the head-length measure- 
ment, which is taken to the posterior edge of the opercle. 
Figure 23 gives the curves for the remainder of central Alaska. The dotted 
curve does not include the Dogfish Bay or Russian Harbor samples. All of the com- 
ponent curves, except that for Halibut Cove in 1927, fall in one group. Here again 
condition is a disturbing element. The 1927 Halibut Cove curve comes considerably 
BODY- LENGTH IN MILLIMETERS 
Figure 21.— Percentage head length is of body length. Showing: A, southeastern Alaska, (1) Craig: 
(2) Stephens Passage; B, Prince William Sound; and C, Shuyak Strait, lower Kachemak Bay, and 
Halibut Cove 
above that for 1926. Reference to the section on condition shows that the 1927 fish 
were much thinner than the 1926 fish. 
The Dogfish Bay curve is 1 per cent higher than the combined curve for Shuyak 
Strait, Kachemak Bay, and Halibut Cove — a difference of 2.5 millimeters in fish of 
250 millimeters in body length. The Dogfish Bay fish, although taken in August, 
were very thin, yet it seems improbable that such a difference could be entirely 
dependent on condition. 
Certain conclusions may be drawn from the head-length analysis: 
1. The head lengths in general decrease from the south toward the north and 
west, thus showing a change with distance, as do the vertebral counts. 
2. While the differences between the curves can not be calculated mathematically 
owing to the differences in slope, the fact that the curves for the individual localities 
